JP7031669B2 - Information processing equipment, information processing methods and programs - Google Patents

Description

The present disclosure relates to information processing devices, information processing methods and programs.

It is known to use an HRV (Heart Rate Variability) index based on heart rate variability when evaluating the degree of psychological stress and the autonomic nervous function. For the HRV index, for example, a heartbeat interval (also called an RR interval (RRI)) is calculated by an electrocardiogram (ECG) obtained by attaching an electrode or the like to a part of the user's body and measuring it. It can be obtained from the calculated RRI. The HRV index can also be obtained from the pulse interval (Pulse Rate Interval: PPI) calculated from the pulse fluctuation having a high correlation with the heart rate variability. Such an acquisition device for RRI or the like is disclosed in the following Patent Documents 1 to 3.

Japanese Unexamined Patent Publication No. 7-284482 Japanese Unexamined Patent Publication No. 2010-162282 Japanese Unexamined Patent Publication No. 2009-261419

In recent years, sensors for detecting heart rate variability and pulse variability have become smaller, and it has become possible for a user to wear the sensor and constantly measure heart rate variability and pulse variability. As a result, the measurement is performed even when the user is freely moving (under free movement) such as daily movements, and the user's state at the time of measuring heart rate variability and pulse variability is at rest and in the same posture. It may not be maintained. That is, it cannot be said that the measurement of heart rate variability and pulse variability is always in a suitable state. Since the heart rate variability and pulse variability measured in such a state may include, for example, noise caused by the movement of the user, the measured heart rate variability and pulse variability are based on these. The reliability of the calculated HRV index and the like may be low. However, even an HRV index or the like relating to a user who can act freely is required to be an index having high reliability.

Therefore, in the present disclosure, we propose a new and improved information processing device, information processing method, and program that can make the measurement suitable for the measurement of heart rate variability or pulse variability in a user under free movement. ..

According to the present disclosure, the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user, or the physical index indicating the physical condition of the user calculated from the pulsation fluctuation data. An information processing apparatus including a reliability calculation unit for calculating reliability and a control unit for controlling various processes based on the calculated reliability is provided.

Further, according to the present disclosure, the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user, or the body indicating the physical condition of the user calculated from the pulsation fluctuation data. An information processing method including calculating the reliability of an index and controlling various processes based on the calculated reliability is provided.

Further, according to the present disclosure, the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user, or the body indicating the physical condition of the user calculated from the pulsation fluctuation data. A program for realizing a function of calculating the reliability of an index and a function of controlling various processes based on the calculated reliability on a computer is provided.

As described above, according to the present disclosure, there is provided an information processing device, an information processing method and a program capable of setting a suitable state for measuring heart rate variability or pulse variability in a user under free movement. be able to.

It should be noted that the above effects are not necessarily limited, and either along with or in place of the above effects, any of the effects shown herein, or any other effect that can be ascertained from this specification. May be played.

It is explanatory drawing explaining the structural example of the information processing system 1 which concerns on embodiment of this disclosure. It is a block diagram which shows the structure of the wearable device 10 which concerns on the same embodiment. It is explanatory drawing explaining the PPG sensor part 122 which concerns on the same embodiment. It is explanatory drawing which shows an example of the pulse wave signal obtained by the PPG sensor unit 122 which concerns on the same embodiment. It is explanatory drawing which shows an example of the time series data of PPI obtained by the PPG sensor unit 122 which concerns on the same embodiment. It is explanatory drawing which shows the wearing example of the wearable device 10 which concerns on the same embodiment. It is a block diagram which shows the structure of the server 30 which concerns on the same embodiment. It is a block diagram which shows the structure of the control part 330 which concerns on the same embodiment. It is explanatory drawing which showed the data flow which concerns on the same embodiment. It is explanatory drawing (the 1) which showed an example of the appearance pattern of the categorized abnormal value which concerns on the same embodiment. It is explanatory drawing (the 2) which showed an example of the appearance pattern of the categorized abnormal value which concerns on the same embodiment. It is explanatory drawing explaining the correction of the abnormal value which concerns on the 1st method of the same embodiment. It is explanatory drawing (the 1) explaining the correction of the abnormal value which concerns on the 2nd method of the same embodiment. It is explanatory drawing (the 2) explaining the correction of the abnormal value which concerns on the 2nd method of the same embodiment. It is explanatory drawing (the 3) explaining the correction of the abnormal value which concerns on the 2nd method of the same embodiment. It is explanatory drawing (the 4) explaining the correction of the abnormal value which concerns on the 2nd method of the same embodiment. It is explanatory drawing (the 5) explaining the correction of the abnormal value which concerns on the 2nd method of the same embodiment. It is explanatory drawing (the 6) explaining the correction of the abnormal value which concerns on the 2nd method of the same embodiment. It is explanatory drawing (the 7) explaining the correction of the abnormal value which concerns on the 2nd method of the same embodiment. It is explanatory drawing (the 8) explaining the correction of the abnormal value which concerns on the 2nd method of the same embodiment. It is explanatory drawing explaining the calculation of the reliability which concerns on the 2nd method of the same embodiment. It is explanatory drawing explaining the calculation of the reliability which concerns on the 3rd method of the same Embodiment. It is explanatory drawing explaining the calculation of the reliability which concerns on the 4th method of the same embodiment. It is explanatory drawing (the 1) explaining the output of the reliability which concerns on the 1st method of the same embodiment. It is explanatory drawing (the 2) explaining the output of the reliability which concerns on the 1st method of the same embodiment. It is explanatory drawing which shows an example of the hardware composition of the information processing apparatus 900 which concerns on this embodiment.

Preferred embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. In the present specification and the drawings, components having substantially the same functional configuration are designated by the same reference numerals, so that duplicate description will be omitted.

Further, in the present specification and the drawings, a plurality of components having substantially the same or similar functional configurations may be distinguished by adding different numbers after the same reference numerals. However, if it is not necessary to distinguish each of the plurality of components having substantially the same or similar functional configurations, only the same reference numerals are given. Further, similar components of different embodiments may be distinguished by adding different alphabets after the same reference numerals. However, if it is not necessary to distinguish each of the similar components, only the same reference numerals are given.

The explanations will be given in the following order.
1. 1. Overview of information processing system 1 according to this embodiment 1.1 Overview of information processing system 1 1.2 Configuration of wearable device 10 1.3 Configuration of server 30 according to this embodiment 2. Creation of embodiment according to this disclosure 3. Detailed configuration of the control unit 330 according to the present embodiment 4. Information processing method according to this embodiment 4.1 Abnormal value detection 4.2 Parameter 4.3 Correction of abnormal value 4.4 Calculation of reliability 4.5 Output of reliability 5. Summary 6. Hardware configuration 7. supplement

<< 1. Outline of information processing system 1 according to this embodiment >>
<1.1 Overview of Information Processing System 1>
Next, the configuration according to the embodiment of the present disclosure will be described. First, the configuration according to the embodiment of the present disclosure will be described with reference to FIG. FIG. 1 is an explanatory diagram illustrating a configuration example of the information processing system 1 according to the present embodiment.

As shown in FIG. 1, the information processing system 1 according to the present embodiment includes a wearable device 10, a server 30, and a user terminal 50, which are communicably connected to each other via a network 70. Specifically, the wearable device 10, the server 30, and the user terminal (output unit) 50 are connected to the network 70 via a base station (for example, a mobile phone base station, a wireless LAN access point, etc.) (not shown). To. As the communication method used in the network 70, any method can be applied regardless of whether it is wired or wireless (for example, WiFi (registered trademark), Bluetooth (registered trademark), etc.), but stable operation is maintained. It is desirable to use a communication method that can be used.

(Wearable device 10)
The wearable device 10 can be a device that can be attached to a part of the user's body (earlobe, neck, arm, wrist, ankle, etc.) or an implant device (implant terminal) inserted into the user's body. More specifically, the wearable device 10 is of various types such as HMD (Head Mounted Display) type, ear device type, anklet type, bracelet type, collar type, eyewear type, pad type, batch type, and garment type. Wearable devices can be adopted. Further, the wearable device 10 is, for example, a PPG (PhotoPlethysmo Graphy) sensor unit (pulse wave sensor) 122 that detects a pulse wave signal (beat fluctuation data) due to the user's pulse, and a motion that detects an exercise state due to the user's movement. Built-in sensors such as the sensor unit 124 (see FIG. 2). The details of the wearable device 10 will be described later.

In the following, it is assumed that the wearable device 10 is provided with the PPG sensor unit 122 (see FIG. 2). However, in the present embodiment, instead of the PPG sensor unit 122, an ECG (Electrocardiogram) sensor (not shown) that detects the user's electrocardiogram via an electrode (not shown) attached to the user's body is provided. May be good. That is, in the embodiment described below, the pulse interval (Pulse Rate Interval: PPI) that can be obtained from the pulse wave signal, or the RR interval that is the heartbeat interval that can be obtained from the electrocardiogram. (RRI) can be used to calculate an HRV (Heart Rate Variability) index, which is a physical index indicating the state of the user's body.

(Server 30)
The server 30 is composed of, for example, a computer or the like. The server 30 processes, for example, the information acquired by the wearable device 10, or transmits the information obtained by the processing to another device (for example, the user terminal 50). The details of the server 30 will be described later.

(User terminal 50)
The user terminal 50 is a terminal for presenting information provided by the present embodiment (for example, time series data of PPI, HRV index, reliability described later, etc.) to the user. For example, the user terminal 50 can be a device such as a tablet PC (Personal Computer), a smartphone, a mobile phone, a laptop PC, a notebook PC, or an HMD.

Note that, in FIG. 1, the information processing system 1 according to the present embodiment is shown as including one wearable device 10 and a user terminal 50, but the present embodiment is not limited to this. .. For example, the information processing system 1 according to the present embodiment may include a plurality of wearable devices 10 and a user terminal 50. Further, the information processing system 1 according to the present embodiment may include, for example, another communication device such as a relay device for transmitting information from the wearable device 10 to the server 30. Further, in the present embodiment, the wearable device 10 may be used as a stand-alone type device. In this case, at least a part of the functions of the server 30 and the user terminal 50 will be performed in the wearable device 10.

<1.2 Configuration of wearable device 10>
Next, the configuration of the wearable device 10 according to the embodiment of the present disclosure will be described with reference to FIGS. 2 to 6. FIG. 2 is a block diagram showing a configuration of the wearable device 10 according to the present embodiment. FIG. 3 is an explanatory diagram illustrating the PPG sensor unit 122 according to the present embodiment. FIG. 4 is an explanatory diagram showing an example of a pulse wave signal obtained by the PPG sensor unit 122 according to the present embodiment. FIG. 5 is an explanatory diagram showing an example of time-series data of PPI obtained by the PPG sensor unit 122 according to the present embodiment. FIG. 6 is an explanatory diagram showing an example of wearing the wearable device 10 according to the present embodiment.

As shown in FIG. 2, the wearable device 10 mainly includes an input unit 100, an output unit 110, a sensor unit 120, a control unit 130, a communication unit 140, and a storage unit 150. The details of each functional part of the wearable device 10 will be described below.

(Input unit 100)
The input unit 100 receives data and commands input to the wearable device 10. More specifically, the input unit 100 is realized by a touch panel, a button, a microphone, a drive, or the like.

(Output unit 110)
The output unit 110 is a device for presenting information to the user, and outputs various information to the user, for example, by image, voice, light, vibration, or the like. The output unit 110 includes a display, a speaker, earphones, and a light emitting element (for example, light).
It is realized by Emitting Diode (LED), vibration module, etc. The function of the output unit 110 may be provided by the user terminal 50 described later.

(Sensor unit 120)
The sensor unit 120 is provided in the wearable device 10 mounted on the user's body, and has a PPG sensor unit 122 that detects the user's pulse wave signal. Further, the sensor unit 120 may include a motion sensor unit 124 for detecting the motion state of the user. The PPG sensor unit 122 and the motion sensor unit 124 included in the sensor unit 120 will be described below.

-PPG sensor unit 122-
The PPG sensor unit 122 is attached to a part of the body such as the user's skin (for example, earlobe, neck, arms, wrists, ankles, etc.) in order to detect the user's pulse wave signal. Here, the pulse wave signal means that the muscles of the heart contract at a constant rhythm (beats, and the number of beats in the heart for a unit time is called the heart rate), so that blood is sent to the whole body through arteries. As a result, a change in pressure occurs on the inner wall of the artery, and it is a waveform due to the pulsation of the artery that appears on the surface of the body or the like. The PPG sensor unit 122 irradiates the blood vessels 202 in the user's measurement site 200 such as the hand, arm, neck, and leg with light in order to acquire the pulse wave signal as shown in FIG. Detects light scattered by moving substances and stationary living tissues. Since the irradiated light is absorbed by the red blood cells in the blood vessel 202, the amount of light absorbed is proportional to the amount of blood flowing through the blood vessel 202 in the measurement site 200. Therefore, it is possible to know the change in the flowing blood volume by detecting the intensity of the scattered light. Furthermore, from the change in blood flow, it is possible to detect the pulsatile waveform of the artery that changes the blood flow, that is, the pulse wave signal. In addition, such a method is called a photoelectric volume pulse wave (Photoplethysmogramhy; PPG) method.

Specifically, the PPG sensor unit 122 incorporates a small laser (not shown) capable of irradiating coherent light, and irradiates light having a predetermined wavelength, for example, around 850 nm. In this embodiment, the wavelength of the light emitted by the PPG sensor unit 122 can be appropriately selected. Further, the PPG sensor unit 122 has a built-in photodiode (PhotoDetector: PD), for example, and acquires a pulse wave signal by converting the detected light intensity into an electric signal. The PPG sensor unit 122 may incorporate a CCD (Charge Coupled Devices) type sensor, a CMOS (Complementary Metal Oxide Sensor) type sensor, or the like instead of the PD. Further, one or a plurality of PDs and the like as described above may be provided in the PPG sensor unit 122.

In this embodiment, the pulse wave signal is not limited to the acquisition by using the above-mentioned PPG method, and the pulse wave signal may be acquired by another method. For example, in the present embodiment, the pulse wave signal may be acquired by the laser Doppler method. In the laser Doppler method, when the laser light is applied to the measurement site 200 of the user, the light is scattered by a scattering substance (mainly red blood cells) moving in the user's blood vessel 202, and the scattered light is the Doppler of blood flow. This is a method that utilizes the effect of causing a frequency shift. Further, in the present embodiment, the pulse wave signal may be detected by using a dynamic light scattering (DLS) method. Similar to the above, the DLS method utilizes the fact that light is scattered by a scattering substance that moves in the blood vessel 202 when irradiated with laser light, and the scattered light causes interference light due to the Doppler effect. Is.

Then, the PPG sensor unit 122 can detect the pulse wave signal as time-series data having a plurality of peaks as shown in FIG. Here, to explain with reference to FIG. 4, the peak interval of a plurality of peaks appearing in the pulse wave signal is referred to as a pulse interval (PPI). The PPI can be acquired by processing the pulse wave signal detected by the PPG sensor unit 122. FIG. 5 shows an example of the time series data of the PPI acquired in this way. Although the value of each PPI fluctuates with time as shown in FIG. 5, it is known that the value is almost normally distributed when the user's condition is stable. By processing the time series data of PPI as shown in FIG. 5, that is, the data group of the PPI value, various HRV (Heart Rate Variability) indexes that are indicators of the physical condition of the user can be calculated. The details of various HRV indexes will be described later.

-Motion sensor unit 124-
The motion sensor unit 124 detects the state of the user by detecting the change in the acceleration generated by the movement of the user. Based on the detected state of the user, the reliability as an index indicating whether the pulse wave signal at that time is appropriately measured is calculated. The details of the reliability will be described later. For example, the motion sensor unit 124 includes an acceleration sensor, a gyro sensor, a geomagnetic sensor, and the like.

Further, the motion sensor unit 124 may be an image pickup device (not shown) that captures a user by using various members such as an image pickup element and a lens for controlling the image formation of a subject image on the image pickup element. good. In this case, the user's actions and the like are captured in the image captured by the image pickup device. Further, the motion sensor unit 124 may include an infrared sensor, an ultrasonic sensor, or the like (not shown) capable of recognizing the user's movement. It should be noted that such an image pickup device, an infrared sensor, and the like may be installed around the user as a device separate from the wearable device 10.

Further, the motion sensor unit 124 may include a positioning sensor (not shown). The positioning sensor is a sensor that detects the position of the user to which the wearable device 10 is attached, and may be specifically a GNSS (Global Navigation Satellite System) receiver or the like. In this case, the positioning sensor can generate sensing data indicating the latitude and longitude of the user's current location based on the signal from the GNSS satellite, and can detect the movement (movement) of the user from the change in the sensing data. For example, since it is possible to detect the relative positional relationship of the user from information such as RFID (Radio Frequency Identification), Wi-Fi access point, and radio base station, such a communication device is used as the positioning sensor. It is also possible to do.

Further, the sensor unit 120 may include various biosensors (not shown). The biosensor is a sensor that detects biometric information indicating the state of the user. The biosensor is, for example, directly or indirectly attached to a part of the user's body and measures one or more sensors for measuring the user's brain waves, respiration, sweating, myoelectric potential, skin temperature, skin electrical resistance, and the like. including. Further, the sensor unit 120 may include various other sensors such as the pressure sensor unit 14 (see FIG. 23). Further, the sensor unit 120 may incorporate a clock mechanism (not shown) for grasping an accurate time, and may associate the acquired pulse wave signal or the like with the time when the pulse wave signal or the like is acquired. In this embodiment, the PPG sensor unit 122 and the motion sensor unit 124, which are the two sensors of the sensor unit 120, may be provided in separate wearable devices 10. By doing so, since the configuration of each wearable device 10 can be made compact, each wearable device 10 can be attached to various parts of the user's body.

(Control unit 130)
The control unit 130 is provided in the wearable device 10 and can control each block of the wearable device 10 or acquire PPI time series data from the pulse wave signal output from the PPG sensor unit 122 described above. can. The control unit 130 is realized by hardware such as a CPU (Central Processing Unit), a ROM (Read Only Memory), and a RAM (Random Access Memory). The function of the control unit 130 may be provided by the server 30 or the user terminal 50, which will be described later.

(Communication unit 140)
The communication unit 140 is provided in the wearable device 10 and can transmit and receive information to and from external devices such as the server 30 and the user terminal 50. In other words, the communication unit 140 can be said to be a communication interface having a function of transmitting and receiving data. The communication unit 140 is realized by a communication device such as a communication antenna, a transmission / reception circuit, and a port.

(Memory unit 150)
The storage unit 150 is provided in the wearable device 10 and stores programs, information, and the like for the control unit 130 described above to execute various processes, and information obtained by the processes. The storage unit 150 is realized by, for example, a non-volatile memory such as a flash memory.

As described above, as the wearable device 10, various types of wearable devices such as eyewear type, ear device type, bracelet type, and HMD type can be adopted. FIG. 6 shows an example of the appearance of the wearable device 10. The wearable device 10 shown in FIG. 6 is a wristwatch-type wearable device worn on the user's wrist.

<1.3 Configuration of server 30 according to this embodiment>
Next, the configuration of the server 30 according to the embodiment of the present disclosure will be described with reference to FIG. 7. FIG. 7 is a block diagram showing a configuration of the server 30 according to the present embodiment. As described above, the server 30 is composed of, for example, a computer or the like. As shown in FIG. 7, the server 30 mainly includes an input unit 300, an output unit 310, a control unit 330, a communication unit 340, and a storage unit 350. The details of each functional unit of the server 30 will be described below.

(Input unit 300)
The input unit 300 receives data and command inputs to the server 30. More specifically, the input unit 300 is realized by a touch panel, a keyboard, or the like.

(Output unit 310)
The output unit 310 is composed of, for example, a display, a speaker, a video output terminal, an audio output terminal, and the like, and outputs various information by an image, an audio, or the like.

(Control unit 330)
The control unit 330 is provided in the server 30 and can control each block of the server 30. Specifically, the control unit 330 performs a detection process for detecting an abnormal value from a pulse wave signal (beat fluctuation data), a correction process for correcting the pulse wave signal, and a pulse wave, which are performed in the server 30. It controls various processes such as a calculation process for calculating an HRV index (body index) from a signal. The control unit 330 is realized by hardware such as a CPU, ROM, and RAM, for example. The control unit 330 may execute a part of the functions of the control unit 130 of the wearable device 10. The details of the control unit 330 will be described later.

(Communication unit 340)
The communication unit 340 is provided in the server 30 and can transmit and receive information to and from an external device such as a wearable device 10 and a user terminal 50. The communication unit 340 is realized by a communication device such as a communication antenna, a transmission / reception circuit, and a port.

(Memory unit 350)
The storage unit 350 is provided in the server 30, and stores programs and the like for the control unit 330 described above to execute various processes, and information obtained by the processes. More specifically, the storage unit 350 can store a database (DB) 352 (see FIG. 8) and the like including time-series data of PPIs acquired from wearable devices 10 worn by a plurality of users. .. The storage unit 350 is realized by, for example, a magnetic recording medium such as a hard disk (Hard Disk: HD), a non-volatile memory, or the like.

<< 2. Technical Background of Embodiments Related to the Disclosure >>
The outline of the information processing system 1, the wearable device 10 and the server 30 included in the information processing system 1 according to the embodiment of the present disclosure has been described above. Next, before explaining the details of the embodiment according to the present disclosure, the technical background of the embodiment according to the present disclosure will be described.

As explained above, the HRV index used when evaluating the degree of psychological stress and the evaluation of autonomic nervous function can be obtained from, for example, RRI calculated from heart rate variability (electrocardiogram). can. Furthermore, the HRV index can also be obtained from PPI calculated from pulse variability having a high correlation with heart rate variability. Since such heart rate variability and pulse variability are affected not only by changes in the user's autonomic nervous system but also by changes in the user's physical condition, the user's state at the time of measurement is maintained at rest and in the same posture. It is desirable that it is done.

By the way, in recent years, sensors for detecting heart rate variability and pulse variability have become smaller, and it has become possible for a user to wear the sensor and constantly measure heart rate variability and pulse variability. As a result, the measurement is performed even when the user is freely moving (under free movement) such as daily movements, and the user's state at the time of measuring heart rate variability and pulse variability is at rest and in the same posture. It cannot be said that it is maintained. Since the heart rate variability and pulse variability measured in such a state may include, for example, noise caused by the movement of the user, the measured heart rate variability and pulse variability are based on these. The reliability of the calculated HRV index and the like may be low.

Therefore, in view of the above situation, the present inventor calculates the reliability for measuring heart rate variability and pulse variability, presents the calculated reliability together with the HRV index to the user, and measures the user's condition. I came up with the idea of inducing an ideal state. For example, when the information processing system 1 according to the embodiment of the present disclosure presents a low reliability, the user is guided to an operation of resting and maintaining the same posture in an attempt to improve the reliability. Will be done. As a result, the user's state approaches the ideal state for the measurement, so that the measurement of the user's heart rate variability or pulse variability can transition to a suitable state. Further, the present inventor feeds back the calculated reliability to the control of the information processing system 1 to calculate a highly accurate HRV index and suppress the battery consumption of the information processing system 1. I also came up with the idea of controlling it appropriately.

That is, in the present disclosure described below, we propose an information processing device, an information processing method, and a program that can make the measurement suitable for the measurement of heart rate variability or pulse variability in a user under free movement.

In addition, heart rate variability and pulse variability data (pulse wave signal) are abnormal due to, for example, "user's body movement", "physical characteristics", "measurement equipment noise", "measurement algorithm error", etc. May contain values (noise). When the HRV index is calculated using the data of the heart rate variability and the pulse fluctuation including such an abnormal value, it may deviate from the correct HRV index to be calculated. Therefore, when calculating the HRV index, it is required to deal with the above-mentioned abnormal value in order to avoid deviation from the correct HRV index to be calculated.

Therefore, in view of the above situation, the present inventor accurately detects an abnormal value (noise) from the data of the heart rate variability and the pulse fluctuation, and corrects the data of the heart rate variability and the pulse fluctuation based on the detected abnormal value. I came up with that. By doing so, it is possible to prevent the calculated HRV index from deviating from the original correct HRV index to be calculated. That is, in the present disclosure described below, an information processing device, an information processing method, and a program capable of improving the accuracy of the HRV index calculated based on the measurement data of the heart rate variability or the pulse variability are proposed. Hereinafter, such embodiments of the present disclosure will be sequentially described in detail.

<< 3. Detailed configuration of the control unit 330 according to this embodiment >>
Hereinafter, the configuration of the control unit 330 according to the present embodiment will be described with reference to FIG. FIG. 8 is a block diagram showing the configuration of the control unit 330 according to the present embodiment. As shown in FIG. 8, the control unit 330 mainly includes a detection unit 332, a detection correction control unit 334, a correction unit 336, a reliability calculation unit 338, and an HRV index calculation unit 342. The details of each functional unit of the control unit 330 will be described below.

(Detection unit 332)
The detection unit 332 detects an abnormal value from the time series data of PPI. Here, the abnormal value means a value that is statistically significantly different from the time series data of PPI, that is, the data group of PPI value due to noise such as an external impact. Further, the detection unit 332 generates a data string in which a flag indicating that the PPI value is estimated to be an abnormal value is added to the PPI value estimated to be an abnormal value in the time series data of the PPI, and the correction unit 336 described later generates a data string. Output. For example, the detection unit 332 assigns "1" to the PPI value estimated to be an abnormal value and "0" to the PPI value estimated to be a normal value. The details of the detection of the abnormal value by the detection unit 332 will be described later.

(Detection correction control unit 334)
The detection correction control unit 334 corrects various parameters used for detecting an abnormal value in the detection unit 332 described above. For example, the detection correction control unit 334 calculates the mean value and standard deviation from the data group of the PPI value acquired from the pulse wave signal of the user measured in the past, and the above-mentioned various parameters are calculated based on the calculated mean value and the like. To update. By doing so, it is possible to detect the abnormal value using a parameter suitable for each user, so that the accuracy of detecting the abnormal value can be improved. The parameter may be updated periodically by the detection / correction control unit 334, or may be performed when the user selects to perform the update. Further, as the parameter, a fixed value determined by conducting experiments and observations in advance may be used. In this case, the detection correction control unit 334 does not update the parameter. The details of updating various parameters by the detection / correction control unit 334 will be described later.

(Correction unit 336)
The correction unit 336 performs correction processing such as interpolation or removal of an abnormal value on the time-series data of the PPI to which the abnormality flag is added obtained from the detection unit 332. The content of the correction process performed by the correction unit 336 can be selected according to the type of the HRV index calculated by the HRV index calculation unit 342, which will be described later. Further, the correction unit 336 may select the content of the correction process according to the expression pattern of the abnormal value. The time-series data of the PPI corrected by the correction unit 336 is output to the HRV index calculation unit 342 described later. The details of the correction process by the correction unit 336 will be described later.

(Reliability calculation unit 338)
The reliability calculation unit 338 determines the time-series data of the PPI obtained from the pulse wave signal and the time-series data of the PPI from the viewpoint of whether or not the user's state is in an appropriate state for the measurement at the time of measuring the pulse wave signal. The reliability of the HRV index obtained from is calculated. Generally, when the HRV index is calculated for the evaluation of the degree of psychological stress or the evaluation of the autonomic nervous function, the user's state at the time of measuring the pulse wave signal which is the basic data of these is a rest and the same posture. It is desirable that it be maintained at. This is because the measured pulse wave signal is affected not only by changes in the user's autonomic nervous system but also by changes in the user's physical condition. Therefore, if the user state at the time of measuring the pulse wave signal is resting and the posture is the same, the reliability of the PPI time series data and the HRV index is high, and the user state at the time of measurement is the above ideal state. The farther away from, the lower the reliability. Therefore, the reliability calculation unit 338 detects the user's state at the time of measuring the pulse wave signal by, for example, the acceleration data of the motion sensor unit 124 of the wearable device 10, and calculates the reliability based on the detected result. .. The reliability calculated by the reliability calculation unit 338 in this way is presented to the user by being output to the user terminal 50 or the like, or is output to the wearable device 10 and used for controlling the wearable device 10. Or something. Further, the reliability is output to another functional unit of the server 30 (for example, the HRV index calculation unit 342 described later), and is used for controlling processing in the other functional unit. Since the reliability is also affected by the mounting state of the PPG sensor unit 122 at the time of measuring the pulse wave signal, it can be calculated based on the mounting state. The details of the reliability calculation by the reliability calculation unit 338 will be described later.

(HRV index calculation unit 342)
The HRV index calculation unit 342 calculates various HRV indexes using the time series data of the PPI corrected by the correction unit 336. The HRV index calculation unit 342 calculates, for example, as an HRV index, RMSD (Root Mean Square Seccess Difference), SDNN (Standard Deviation of the Normal to Normal, etc.), L / F. Further, these calculated HRV indexes can be further processed and used for the evaluation of the sleep state, the evaluation of the mental stress degree, the evaluation of the relaxation degree, the evaluation of the concentration degree, and the like. The HRV index or the like calculated by the HRV index calculation unit 342 in this way is output to the user terminal 50 or the like and presented to the user, for example.

For example, RMSSD is the root mean square of the squares of the differences between adjacent PPI values in a time series. The RMSSD is considered to be an index showing the tension state of the vagus nerve, which is one of the cranial nerves.

For example, SDNN is the standard deviation of a data set of PPI values within a predetermined period (eg, 120 seconds). The SDNN is considered to be an index showing the activity status of the autonomic nervous system including both the sympathetic nerve and the parasympathetic nerve.

For example, LF / HF is the ratio of the power spectrum of the low frequency component (eg 0.004 to 0.15 Hz) to the power spectrum of the high frequency component (eg 0.15 to 0.4 Hz) of the time series data of PPI. be. LF / HF is considered to be an index showing the balance between the sympathetic nerve and the parasympathetic nerve. When the LF / HF is high, the sympathetic nerve is dominant, and when the LF / HF is low, the parasympathetic nerve is dominant. It is thought to indicate a certain state.

<< 4. Information processing method according to this embodiment >>
The detailed configuration of the control unit 330 according to the present embodiment has been described above. Next, the outline of the information processing method according to the present embodiment will be described with reference to FIG. FIG. 9 is an explanatory diagram showing a data flow according to the present embodiment.

As shown in FIG. 9, the pulse wave signal detected by the PPG sensor unit 122 provided in the wearable device 10 is transmitted by the control unit 130 of the wearable device 10 (or may be the control unit 330 of the server 30). Process and get PPI time series data. Further, the time-series data of the PPI is processed by the detection unit 332 of the server 30, and an abnormal value included in the time-series data of the PPI is detected (abnormal value detection). Then, based on the detected abnormal value, the correction unit 336 corrects the time-series data of the PPI, and outputs the time-series data of the PPI with the corrected abnormal value to the HRV index calculation unit 342 (correction of the abnormal value). At this time, the reliability calculation unit 338 of the server 30 calculates the reliability by using the acceleration data of the motion sensor unit 124 of the wearable device 10 (calculation of reliability). Further, the HRV index calculation unit 342 calculates the HRV index based on the time-series data of the PPI in which the abnormal value is corrected, and outputs the calculated HRV index together with the reliability to the user terminal 50 or the like (output of the reliability). ). The details of the processing at each stage of the information processing method according to the present embodiment will be described below.

<4.1 Abnormal value detection>
In the detection of the abnormal value according to the present embodiment described below, the continuous PPI value for a predetermined length period is extracted from the acquired PPI time series data, and the extracted PPI time series data is categorized as an abnormality. Outliers are detected by comparing with the expression pattern of the values. That is, according to the present embodiment, it is not determined whether or not the PPI value is an abnormal value for each PPI value included in the time series data of the PPI, but a plurality of PPI values are determined at one time, so that the abnormal value is detected. It is possible to reduce the processing and processing time related to the above.

Hereinafter, an example of detecting an abnormal value according to the present embodiment will be described with reference to FIGS. 10 and 11. 10 and 11 are explanatory views showing an example of the appearance pattern of the categorized abnormal value according to the present embodiment. Experiments and observations were conducted in advance, and a period of a predetermined length of time-series data of PPI determined to contain an abnormal value based on the observation result was extracted, and the behavior of the extracted period was categorized. The expression patterns of the eight outliers are shown in FIGS. 10 and 11. Specifically, in the expression pattern of each abnormal value having a predetermined length period, five consecutive PPI values (indicated by black dots in the figure) are included, and further, it is determined to be an abnormal value. PPI values are indicated by circles in the figure. In the embodiment described below, whether or not the newly acquired time-series data of the PPI corresponds to the expression pattern of such an abnormal value is determined based on the magnitude relationship of the numerical values, thereby determining the abnormal value. Is detected.

First, the time-series data of the PPI for a predetermined length period including the five PPI values from the beginning of the acquired time-series data of the PPI is extracted. In the following description, the five PPI values included in the extracted PPI time-series data are _pn , _{pn + 1} , _{pn + 2} , _{pn + 3} , and _{pn + 4} , respectively, in chronological order.

Next, in the detection of the abnormal value according to the present embodiment, the following mathematical formulas (1) to (5) are used to use the following parameters loc, th _roc1 , th _roc2 , th _eto1 and th for detecting the abnormal value. _{Calculate eo2} . The th _roc1 , th _roc2 , _theto1 , and _theto2 are obtained by statistically processing the time-series data of a plurality of PPIs acquired by measuring the pulse wave signals a plurality of times for various users in advance. It is calculated based on the mean value μ'and the standard deviation σ'obtained by statistically processing the time-series data of the obtained mean value μ and the standard deviation σ and the difference value of the PPI values adjacent to each other. Further, in the following description, it is assumed that th _roc1 , th _roc2 , th _eo1 , and th _oto2 are predetermined fixed values.

It should be noted that α included in the mathematical formulas (2) to (5) is a value determined in advance based on experiments, observations, etc., and can be set to 1.0, for example.

(Case 1)
When the following mathematical formulas (6) and (7) are satisfied in _pn and _{pn + 1} included in the time-series data of the extracted PPI, the time-series data of the extracted PPI is the case of FIG. It is determined that it corresponds to 1. In this case, _{pn + 1} and _{pn + 2} are detected as abnormal values.

(Case 2)
When the following mathematical formulas (8) and (9) are satisfied in _pn and _{pn + 1} included in the time-series data of the extracted PPI, the time-series data of the extracted PPI is the case of FIG. It is determined that it corresponds to 2. In this case, _{pn + 1} is detected as an abnormal value.

(Case 3)
When the following mathematical formulas (10) and (11) are satisfied in _pn and _{pn + 1} included in the time-series data of the extracted PPI, the time-series data of the extracted PPI is the case of FIG. It is determined that it corresponds to 3. In this case, _{pn + 1} and _{pn + 2} are detected as abnormal values.

(Case 5)
When the following mathematical formulas (12) and (13) are satisfied in _pn and _{pn + 1} included in the time-series data of the extracted PPI, the time-series data of the extracted PPI is the case of FIG. It is determined that it corresponds to 5. In this case, _{pn + 1} and _{pn + 2} are detected as abnormal values.

(Case 6)
When the following mathematical formulas (14) and (15) are satisfied in _pn and _{pn + 1} included in the time-series data of the extracted PPI, the time-series data of the extracted PPI is the case of FIG. It is determined that it corresponds to 6. In this case, _{pn + 1} is detected as an abnormal value.

(Case 7)
When the following mathematical formulas (16) and (17) are satisfied in _pn and _{pn + 1} included in the time-series data of the extracted PPI, the time-series data of the extracted PPI is the case of FIG. It is determined that it corresponds to 7. In this case, _{pn + 1} and _{pn + 2} are detected as abnormal values.

(Case 4)
Next, in cases 1 to 3 and cases 5 to 7 described above, _pn , _{pn + 1} , _{pn + 2} , _{pn + 3, and pn + 4} included in the extracted PPI time series data are as follows. When the mathematical formulas (18) to (20) of the above are satisfied, it is determined that the extracted time series data of PPI corresponds to the case 4 of FIG. In this case, _pn , _{pn + 1} , _{pn + 2} , and _{pn + 3} are detected as abnormal values.

(Case 8)
When the following mathematical formulas (21) to (23) are satisfied in _pn , _{pn + 1} , _{pn + 2} , _{pn + 3, and pn + 4} included in the time series data of the extracted PPI, the time of the extracted PPI. The series data is determined to fall under case 8 of FIG. In this case, _pn , _{pn + 1} , _{pn + 2} , and _{pn + 3} are detected as abnormal values.

A flag indicating that the abnormal value is detected is added to the abnormal value detected as described above. Further, the time-series data of the PPI for a predetermined length period including the following five PPI values is extracted from the acquired time-series data of the PPI, and the abnormal value as described above is detected. Then, the detection of the above-mentioned abnormal value is repeated until the processing for the PPI value at the end point of the acquired PPI time-series data is completed.

In the detection of the abnormal value described above, it is determined whether the time-series data of the extracted PPI for a predetermined length period corresponds to any of the eight abnormal value expression patterns shown in FIGS. 10 and 11. However, the present embodiment is not limited to such a method. For example, in the present embodiment, it is determined whether or not the time-series data of the extracted PPI for a predetermined length period corresponds to the expression pattern of four abnormal values instead of eight, and the abnormal values are detected. You may.

<4.2 parameters>
In the above-mentioned detection of abnormal values, the parameters th _roc1 , th _roc2 , the _too1 , and _theto2 used for detecting the abnormal values have been described as being predetermined fixed values. However, in the present embodiment, the parameters th _roc1 , th _roc2 , the _too1 , and _theto2 are not limited to fixed values, but may be dynamically changing values.

Specifically, the tendency of the pulse wave signal of the user differs according to the physical characteristics of the user, and further changes according to the time of measurement, the age of the user, and the physical condition. Therefore, in order to detect abnormal values with high accuracy, it is preferable to use the parameters th _roc1 , th _roc2 , _theto1 and _theto2 that reflect the influence of the physical characteristics of the user. Therefore, in the present embodiment, when newly measuring the pulse wave signal of the user, for example, the time series data of PPI obtained from the pulse wave signal of the user for one day before the measurement (for example, the figure). The mean values μ and μ'and the standard deviations σ and σ'are calculated using the PPI time-series data stored in the DB352a of 8. Then, in the present embodiment, based on the calculated mean values μ and μ'and the standard deviations σ and σ', the parameters th _roc1 , th _roc2 the _too1 and the too2 are used using the above mathematical formulas (2) to (5) _. Is calculated, and the parameters th _roc1 , th _roc2 , th _eo1 , and the _tho2 are used when detecting an abnormal value. This is because the time-series data of the PPI obtained from the pulse wave signal of the user for one day can be regarded as the data reflecting the physical characteristics and the physical condition of the user. In addition, in order to reflect the current physical condition of the user, the parameters th _roc1 , th _roc2 , the _too1 , and _theto2 may be updated every time a new pulse wave signal is measured or periodically. preferable.

Further, when calculating the mean values μ and μ ′ and the standard deviations σ and σ ′, the PPI time series data in the section with high reliability described later may be used. In addition, when calculating the mean values μ, μ'and standard deviations σ, σ', the contribution of the latest PPI to the mean values μ, μ'and standard deviations σ, σ'is high. In addition, the time-series data of the PPI used for the calculation may be weighted. In addition, when calculating the parameters th _roc1 , th _roc2 , _theto1 , and theto2, based on the sensing data acquired from various _biosensors (not shown) provided in the sensor unit 120 of the wearable device 10. Α included in the formulas (2) to (5) may be changed. That is, in the present embodiment, if the parameters th _{roc1, th roc2, the too1, and theto2 that can accurately detect the abnormal value can be acquired, the parameters th roc1 , th roc2 , and th .} The calculation method of _eto1 and _theto2 is not particularly limited.

<4.3 Correction of abnormal values>
In the correction of the abnormal value according to the present embodiment described below, the abnormal value is detected from the acquired time-series data of the PPI, and the time-series data of the PPI is corrected based on the detection result. According to the present embodiment, since the abnormal value is corrected, the accuracy of various HRV indexes acquired from the time-series data of the corrected PPI can be further improved. In the following, as correction of the abnormal value, mainly two examples of a method of performing correction according to the type of the HRV index and a method of performing correction according to the expression pattern of the abnormal value will be described.

(First method)
First, as an example of correction of an abnormal value, a first method of performing correction according to the type of HRV index will be described with reference to FIG. FIG. 12 is an explanatory diagram illustrating correction of an abnormal value according to the first method of the present embodiment. In detail, the upper part of FIG. 12 shows a correction example of Case A that linearly interpolates the section in which an abnormal value is detected, and the middle part of FIG. 12 shows a correction example of Case B that removes an abnormal value. In the lower part of FIG. 12, a correction example of the case C in which the correction is not performed is shown. According to the first method, since the correction is performed according to the type of the HRV index, the HRV index can be calculated using the time series data of the PPI corrected by the suitable correction method, and the calculated HRV index can be calculated. The accuracy of can be further improved.

Specifically, in this method, when calculating RMSSD as an HRV index, case A shown in the upper part of FIG. 12 is selected, and the section in which an abnormal value is detected is linearly interpolated. Further, in this method, when calculating SDNN as an HRV index, case B shown in the middle of FIG. 12 is selected and an abnormal value is removed. Further, in this method, when calculating LF / HF as an HRV index, case C shown in the lower part of FIG. 12 is selected and no correction is performed. In this case, when calculating the LF / HF, the LF / HF is calculated using the time-series data of the PPI in the longest section in which the abnormal value does not exist continuously.

(Second method)
Next, as an example of the correction of the abnormal value, a second method of performing the correction according to the expression pattern of the abnormal value will be described with reference to FIGS. 13 to 20. 13 to 20 are explanatory views illustrating correction of abnormal values according to the second method of the present embodiment. In this method, the correction method is selected in consideration of the reason for the occurrence of the abnormal value expression pattern (cases 1 to 8) shown in FIGS. 10 and 11. According to the second method, since the correction is performed according to the expression pattern of the abnormal value, the abnormal value can be corrected after considering the reason for the occurrence of the abnormal value. Therefore, the correction of the abnormal value is more preferable. It can be performed.

-Case 1-
First, with reference to FIG. 13, the correction in the case of the case 1 of the appearance pattern of the abnormal value will be described. Case 1 is shown in the middle part of FIG. 13 because, as shown in the upper part of FIG. 13, the correct peak T that should be originally detected was not detected in the acquired pulse wave signal, but the erroneous peak E was detected. This is the case when the time series data of such PPI is acquired. Specifically, in the time series data of PPI shown in the middle of FIG. 13, _{pn + 1} is a large value and _{pn + 2} is a small value due to the detection of an erroneous peak E. That is, _{pn + 1} and _{pn + 2} correspond to abnormal values. The detection of an erroneous peak E is caused by the occurrence of an erroneous peak E in the vicinity of the position where the correct peak T that should be originally detected should appear due to the DC shift of the pulse wave signal. In this case, if the correct peak T is detected, the original _{pn + 1} has a smaller value than the _{pn + 1} in the middle of FIG. 13, and the original _{pn + 2} is pn ₊ 2 shown in the middle of FIG. It is presumed that the value is larger than that. Therefore, in Case 1, as shown in the lower part of FIG. 13, Δβ is calculated according to the following mathematical formula (24), the abnormal value _{pn + 1} is reduced by Δβ, and the abnormal value _{pn + 2} is increased by Δβ. Make corrections. In this way, the abnormal value caused by the DC shift can be corrected.

It should be noted that α included in the mathematical formula (24) is a value determined in advance based on experiments, observations, etc., and can be set to 1.0, for example. Further, the α included in the mathematical formula (22) may be the α used when calculating th _roc1 , th _roc 2, and the like.

-Case 2-
The correction in the case of case 2 of the abnormal value appearance pattern will be described with reference to FIG. In case 2, as shown in the upper part of FIG. 14, one peak T that should be originally detected was not detected in the acquired pulse wave signal, so that the time series of PPI as shown in the middle part of FIG. This is the case when the data is acquired. Specifically, in the time series data of PPI shown in the middle of FIG. 14, one peak T that should be originally detected was not detected, so that there should be two peaks, but two. An erroneous outlier _{pn + 1} with the sum of the peaks appears. It is presumed that the expression of the abnormal value _{pn + 1} was caused by the influence of the user's arrhythmia. Therefore, if the correct peak T is detected, it is presumed that _{pn + 1} is divided into two in the time series data of PPI. Therefore, in Case 2, as shown in the lower part of FIG. 14, p'n _{+ 1} and p'n _{+ 1} are inserted after the abnormal value _pn + 1, and correction is performed to divide _{pn + 1} into two. p'n _{+ 1} and p'n _{+ 1} can be calculated based on the following mathematical formulas (25) and (26). In this way, it is possible to correct an abnormal value caused by a user's arrhythmia or the like.

The α included in the mathematical formula (25) and the mathematical formula (26) is a value determined in advance based on experiments, observations, etc., and can be set to 0.5, for example.

-Case 3-
With reference to FIG. 15, the correction in the case of the case 3 of the appearance pattern of the abnormal value will be described. In case 3, as shown in the upper part of FIG. 15, three peaks should be detected in the acquired pulse wave signal, but two peaks T could not be detected due to one large peak E. This is the case where the time series data of PPI as shown in the middle of FIG. 15 is acquired. Specifically, in this case, as shown in the upper part of FIG. 15, an erroneous peak E due to impact noise or the like is detected, and two correct peaks T that should be originally detected located in the vicinity of the erroneous peak E are detected. It becomes impossible to detect, and as shown in the middle part of FIG. 15, abnormal values _{pn + 1} and _{pn + 2} having large values are detected. In this case, if the correct peak T is detected, it is presumed that the abnormal values _{pn + 1} and _{pn + 2} are divided into three. Therefore, in Case 3, as shown in the lower part of FIG. 15, the abnormal value _{pn + 1} is reduced by Δβ ₁ minute, the abnormal value _{pn + 2} is reduced by Δβ ₂ minutes, and the abnormal value _{pn + 1} and _{pn + 2} are in the middle. _, P'n _{+ 1} is inserted. p'n _{+ 1} can be calculated according to the following mathematical formula (27). In addition, Δβ ₁ and Δβ ₂ can be determined by a value or a mathematical formula determined in advance based on experiments, observations, and the like. In this way, it is possible to correct an abnormal value caused by impact noise or the like.

-Case 4-
With reference to FIG. 16, the correction in the case of the case 4 of the appearance pattern of the abnormal value will be described. In case 4, as shown in the upper part of FIG. 16, the PPI value in the time series data of PPI gradually decreases ( _pn , _{pn + 1} , _{pn + 2} , _{pn + 3} ), and the original detection is performed at a certain point. It has returned to the power value ( _{pn + 4} ). Such abnormal values _pn , _{pn + 1} , _{pn + 2} , and _{pn + 3} detect the original pulse by detecting an erroneous peak generated due to the addition of high frequency noise due to the heartbeat to the pulse wave signal. It is caused by detecting peaks at intervals shorter than the wave intervals. For example, in the PPG sensor unit 122, high frequency noise of about 3 Hz is added to a pulse wave that oscillates with a period of about 1 Hz. In such a case, if the correct peaks are detected, the number of peaks to be detected should be less than the number of 4 peaks detected as outliers. Therefore, in Case 4, as shown in the lower part of FIG. 16, the abnormal values _pn , _{pn + 1} , and _{pn + 2} are corrected to m, and _{pn + 3} is deleted. m can be calculated according to the following mathematical formula (28). In this way, it is possible to correct an abnormal value caused by high frequency noise or the like.

-Case 5-
With reference to FIG. 17, the correction in the case of the case 5 of the appearance pattern of the abnormal value will be described. In case 5, as in case 1, as shown in the upper part of FIG. 17, the correct peak T that should be originally detected was not detected in the acquired pulse wave signal, but the erroneous peak E was detected. This is the case where the PPI time series data as shown in the middle of FIG. 17 is acquired. Specifically, in the time series data of PPI shown in the lower part of FIG. 17, _{pn + 1} is a small value and _{pn + 2} is a large value due to the detection of an erroneous peak E. That is, _{pn + 1} and _{pn + 2} correspond to abnormal values. The detection of an erroneous peak E is caused by the occurrence of an erroneous peak E in the vicinity of the position where the correct peak T that should be originally detected should appear due to the DC shift of the pulse wave signal. In this case, if the correct peak T is detected, the original _{pn + 1} has a larger value than the _{pn + 1} shown in the middle of FIG. 17, and the original _{pn + 2} is shown in the middle of FIG. It is presumed that the value is smaller than that of _{pn + 2} . Therefore, in Case 5, as shown in the lower part of FIG. 17, Δβ is calculated according to the following mathematical formula (29), the abnormal value _{pn + 1} is increased by Δβ, and the abnormal value _{pn + 2} is decreased by Δβ. Make corrections. In this way, the abnormal value caused by the DC shift can be corrected.

It should be noted that α included in the mathematical formula (29) is a value determined in advance based on experiments, observations, etc., and can be set to 1.0, for example.

-Case 6-
With reference to FIG. 18, the correction in the case of the case 6 of the appearance pattern of the abnormal value will be described. In case 6, as shown in the upper part of FIG. 18, one peak should be detected in the acquired pulse wave signal, but an erroneous peak E due to impact noise due to an external impact is also detected. , The case where the time series data of PPI as shown in the middle of FIG. 18 is acquired. In this case, if the correct peak is detected, it is estimated that one peak is detected. Therefore, in Case 6, as shown in the lower part of FIG. 18, a new p'n _{+ 2} is calculated by adding _{pn + 1} to _{pn + 2} according to the following mathematical formula (30). Then, in case 6, _{pn + 2} is corrected to the calculated p'n + 2, and _{pn + 1} is deleted. In this way, it is possible to correct an abnormal value caused by impact noise or the like.

-Case 7-
With reference to FIG. 19, the correction in the case of the case 7 of the appearance pattern of the abnormal value will be described. In case 7, as shown in the upper part of FIG. 19, one peak should be detected in the acquired pulse wave signal, but an erroneous peak E due to impact noise due to an external impact is detected. , This is the case where the PPI series data as shown in the middle of FIG. 19 is acquired. In this case, if the correct peak T is detected, it is estimated that one peak is detected. Therefore, in Case 7, as shown in the lower part of FIG. 19, a new p'n _{+ 1} is calculated by adding _{pn + 1} to _{pn + 2} according to the following mathematical formula (31). Then, in case 7, _{pn + 1} is corrected to the calculated p'n + 1, and _{pn + 2} is deleted. In this way, it is possible to correct an abnormal value caused by impact noise or the like.

-Case 8-
With reference to FIG. 20, the correction in the case of the case 8 of the appearance pattern of the abnormal value will be described. In case 8, as shown in the upper part of FIG. 20, the PPI value in the time series data of PPI gradually increases ( _pn , _{pn + 1} , _{pn + 2} , _{pn + 3} ), and the original detection is performed at a certain point. It has returned to the power value ( _{pn + 4} ). Such abnormal values _pn , _{pn + 1} , _{pn + 2} , and _{pn + 3} occur because the peak that should be detected could not be detected due to the high frequency noise added to the pulse wave signal due to the heartbeat. .. In this case, if the correct peak T is detected, the number of peaks to be detected should be larger than the number of 4 peaks detected as abnormal values. Therefore, in Case 8, as shown in the lower part of FIG. 20, the abnormal values _pn , _{pn + 1} , _{pn + 2} , and _{pn + 3} are replaced with m, and p'n _{+ 3} having the same value of m is _{pn + 3} . Make corrections to be added after. m can be calculated according to the following mathematical formula (32). In this way, it is possible to correct an abnormal value caused by high frequency noise or the like.

In the present embodiment, the correction of the abnormal value is not limited to the above method, and is not particularly limited. In the present embodiment, for example, the abnormal value may be corrected according to the reliability described later. In this case, specifically, the abnormal value detected in the time-series data section of the highly reliable PPI is presumed to have been caused by some change in the user's autonomic nervous system, and is positively abnormal. Choose not to correct the value. On the other hand, for the abnormal values detected in the time series data section of PPI with low reliability, it is estimated that they were caused by external influences such as impact, and the abnormal values should be positively corrected. select. Further, in the present embodiment, for the detection of the abnormal value and the correction of the abnormal value, the tendency of the time series data of the past PPI of the user is learned by machine learning, and the tendency of each user is determined based on the learning result. The abnormal value may be detected and corrected.

<4.4 Calculation of reliability>
In the calculation of the reliability according to the present embodiment described below, when the PPI obtained from the pulse wave signal is obtained from the viewpoint of whether the user's state is in an appropriate state for the measurement at the time of measuring the pulse wave signal. The reliability of the HRV index obtained from the series data and the time series data of the PPI is calculated. As described above, when the HRV index is calculated for the evaluation of the degree of psychological stress and the evaluation of the autonomic nervous function, the user's state at the time of measuring the pulse wave signal which is the basic data of these is resting and It is desirable that the posture is maintained in the same position. Further, in the calculation of the reliability, it can be calculated from the viewpoint of whether or not the mounting state of the PPG sensor unit 122 that detects the pulse wave signal at the time of measuring the pulse wave signal is in an appropriate state. Therefore, in the present embodiment, by calculating the reliability from the above-mentioned viewpoint and presenting the calculated reliability to the user, the user can be made to recognize the reliability of the HRV index presented together. Further, in the present embodiment, the sensor unit 120 of the wearable device 10 can be controlled or the HRV index calculation process can be controlled based on the calculated reliability. Hereinafter, various examples of calculating the reliability according to the present embodiment will be described.

(First method)
First, as an example of the method of calculating the reliability _ri , a first method of calculating the reliability _ri will be described using the sensing data acquired by the motion sensor unit 124 provided in the wearable device 10. Since the sensing data acquired by the motion sensor unit 124 is data that reflects the user's movement, by analyzing such data, the user is in an appropriate state when measuring the pulse wave signal. It can be determined whether or not.

Specifically, in this method, when measuring the pulse wave signal of the user, the acceleration data due to the operation of the user is acquired from the motion sensor unit 124 mounted on the user. Further, in this method, the acceleration norm is calculated from the acquired acceleration data at each timing of sampling the pulse wave signal. Then, in this method, a plurality of calculated acceleration norms are averaged, for example, every second, and a plurality of vectors Ai having the average value obtained by averaging as an element are obtained.

Next, in this method, the mean value μ _i and the standard deviation σ _i are calculated using the values of a plurality of vectors Ai. Further, in this method, the calculated mean value μ _i and the standard deviation σ _i are used to indicate the “rest score” Sr indicating the degree of rest of the user and the small change in the posture of the user, that is, the exercise state. The "posture score" Sp is calculated according to the following mathematical formulas (33) and (34). It is estimated that Sr is a small value if the user is in a resting state, and Sp is a small value if the posture of the user does not change.

In addition, σ _μ and σ _σ in the above equations (33) and (34) are the mean value μi group and the standard deviation calculated in advance from a plurality of vectors Ai obtained from the acceleration data in the daily life of various users. Each standard deviation of the σi group.

It is preferable that the rest score Sr and the posture score Sp are clipped so that these values fall within the range of 0 or more and 1 or less. Then, in this method, the reliability _ri is calculated according to the following mathematical formula (35) by using the calculated rest score Sr and the posture score Sp.

It should be noted that α included in the mathematical formula (33) is a value determined in advance based on experiments, observations, etc., and a numerical value of 0 or more and 1 or less can be set.

The reliability ri thus obtained is presented to, for example, the user. At this time, the reliability ri may be presented to the user together with the calculated rest score Sr and posture score Sp.

Further, in this method, it suffices if the rest score Sr indicating the degree of rest of the user and the posture score _Sp indicating that the change in the posture of the user is small can be calculated. It may be calculated using a statistical value other than the deviation σ _i . Further, in this method, for example, the rest score Sr and the posture score Sp are obtained by using the acceleration data in each axis direction of the XYZ axes acquired from the 3-axis acceleration sensor (not shown) built in the motion sensor unit 124. May be calculated. In this case, the reliability ri in each axial direction may be obtained using the acceleration data in each axial direction as described above, and the linear sum of the obtained reliability ri may be used as the final reliability.

Further, in this method, the posture and state of the user are estimated from the sensing data acquired by the motion sensor unit 124 using an existing algorithm, and the rest score Sr and the posture score Sp are calculated based on the estimation result. You may. More specifically, based on the sensing data acquired by the motion sensor unit 124, it is estimated which state the user is in, such as "sitting", "standing", "walking", and "running". Further, the state of the user and the value of the rest score Sr and the value of the posture score Sp are associated in advance (for example, when the user is in the "sitting position", Sr is 1.0 and the user is "walking". Sr is 0.5 in the case of the state, and Sr is 0.0 in the case of the user in the "running" state), the value of the rest score Sr and the value of the rest score Sr based on the above estimation result. The value of the posture score Sp is selected, and the reliability ri is calculated using the selected value. Further, in this method, when the user's pulse wave signal is assumed, the user's state is estimated from the sensing data acquired by the motion sensor unit 124 using an existing algorithm, and the user's state is estimated within the measurement period of the pulse wave signal. The percentage of time in the longest posture may be the rest score Sr.

(Second method)
Next, as an example of the method of calculating the reliability _ri , a second method of calculating the reliability _ri using the pulse wave signal acquired by the PPG sensor unit 122 provided in the wearable device 10 is shown in FIG. It will be described with reference to 21. FIG. 21 is an explanatory diagram illustrating the calculation of the reliability according to the second method of the present embodiment. According to this method, it is possible to acquire the rest score Sr indicating the rest degree of the user without providing the motion sensor unit 124.

Specifically, the rest score Sr, which indicates the degree of rest of the user, can be calculated by quantifying the fluctuation of the DC component (DC component) contained in the pulse wave signal in a predetermined period. Therefore, in this method, the change in the DC component of the pulse wave signal is detected, the above-mentioned rest score Sr is calculated based on the detection result, and the calculated rest score Sr is used to obtain the reliability as in the first method. Calculate r _i . More specifically, the pulse wave signal includes a pulsatile component (AC component) corresponding to a change in blood flow due to the pulsation of the user's heart, and reflected light from a blood layer other than the pulsation or a tissue other than the blood. Includes a DC component (DC component) corresponding to scattered light. In the case of a non-resting state such as when the user is exercising, external light from other than the PPG sensor unit 122 is emitted due to a change in blood flow due to the user's movement or a misalignment of the PPG sensor unit 122. Upon detection, the DC component of the pulse wave signal fluctuates. For example, in FIG. 21, when the wearable device 10 is attached to the user's arm, that is, the PPG sensor unit 122 is attached to the user's arm and the user's arm is maintained at the height of the user's chest, the wearable device 10 is attached to the user's chest. The fluctuation of the pulse wave signal is shown when the arm is raised and the arm is lowered below the chest. As can be seen from FIG. 21, by raising and lowering the arm on which the PPG sensor unit 122 is attached, fluctuations in the DC component of the pulse wave signal appear as fluctuations in the height of the pulse wave signal in the figure. That is, if the posture of the user is maintained in a constant state, the DC component of the pulse wave signal is maintained at a constant value. As described above, the DC component of the pulse wave signal can be considered as an index reflecting the rest level of the user. Therefore, in this method, for example, the variation of the DC component of the pulse wave signal during the measurement period of the pulse wave signal is statistically processed to calculate the variance (σ ² ), and the reciprocal of the calculated variance is used as the above-mentioned rest score Sr. Can be used.

(Third method)
Next, as an example of the method of calculating the reliability r _i , the third reliability r _i is calculated by using the change in the pulse based on the pulse wave signal acquired by the PPG sensor unit 122 provided in the wearable device 10. The method of the above will be described with reference to FIG. 22. FIG. 22 is an explanatory diagram illustrating the calculation of the reliability according to the third method of the present embodiment. According to this method, it is possible to acquire a posture score Sp indicating that the change in the posture of the user is small without providing the motion sensor unit 124.

Since the pulse wave signal acquired by the PPG sensor unit 122 is data that reflects the change due to the user's pulse, the user's pulse rate at a predetermined time can be obtained by analyzing the pulse wave signal (for example,). , The pulse rate per minute. The pulse rate is also referred to as a beat per minute (bpm)). When the user's state changes, the pulse rate changes accordingly. For example, as shown in FIG. 22, when the user's state changes from a lying lying position to a standing standing state, a change occurs in the pulse rate. More specifically, in the recumbent position, the pulse rate is stable at about 60 to 70 bpm, and in the standing position, the pulse rate is changed to about 80 to 100 bpm. From this, the change in pulse rate can be considered as an index that reflects the small change in the posture of the user. Therefore, in this method, for example, the variation of the pulse rate during the measurement period of the pulse wave signal is statistically processed to calculate the variance (σ ² ), and the reciprocal of the calculated variance can be used as the above-mentioned posture score Sp. .. In this method, the posture score Sp may be calculated using the heart rate instead of the pulse rate.

(Fourth method)
As described above, the reliability ri can also be calculated from the viewpoint of whether or not the _PPG sensor unit 122 that detects the pulse wave signal at the time of measuring the pulse wave signal is in an appropriate state. Therefore, as an example of the method of calculating the reliability _ri , a fourth method of calculating the reliability _ri based on the mounting state of the PPG sensor unit 122 will be described with reference to FIG. 23. FIG. 23 is an explanatory diagram illustrating the calculation of the reliability according to the fourth method of the present embodiment, and in detail, shows the form of the wearable device 10a according to the present method. According to this method, the reliability _ri can be calculated in consideration of the mounting state of the PPG sensor unit 122.

In order to properly acquire the pulse wave signal, the PPG sensor unit 122 is required to be attached to a part of the user's body with an appropriate tightening condition. Therefore, in this method, as shown in FIG. 23, a pressure sensor unit 14 for detecting the mounted state of the PPG sensor unit 122 is provided together with the PPG sensor unit 122 inside the band unit 12 of the wristwatch-type wearable device 10a. .. The pressure sensor unit 14 can detect the pressure on the PPG sensor unit 122 at the time of measuring the pulse wave signal, and can acquire the mounted state of the PPG sensor unit 122 by using the detection result. Then, in this method, the difference between the pressure when ideally mounted and the pressure detected by the pressure sensor unit 14 within the pulse wave signal measurement period is used as an index indicating the mounted state of the PPG sensor unit 122. It is processed mathematically (for example, the dispersion (σ ² ) is calculated), and the reliability _ri is calculated based on the processing result.

Further, in this method, the mounting state of the PPG sensor unit 122 is not limited to being detected by the pressure sensor unit 14 described above, and may be detected by another sensor. For example, in order to properly acquire a pulse wave signal, the PPG sensor unit 122 is required to be in close contact with a part of the user's body. Therefore, a temperature sensor unit (not shown) may be provided inside the band portion 12 of the wristwatch-type wearable device 10a at a location in close contact with the user's skin. The temperature sensor unit may detect the skin temperature of the user, and the detection result may be used to detect the wearing state of the PPG sensor unit 122.

(Fifth method)
Further, the reliability _ri can also be calculated by using the above-mentioned detection of abnormal values. Therefore, as an example of the method of calculating the reliability _ri , a fifth method of calculating the reliability _ri using the detection of an abnormal value will be described. When an abnormal value is detected, it is presumed that the user is not in an appropriate state when measuring the pulse wave signal, or that the PPG sensor unit 122 is not in an appropriate state when measuring the pulse wave signal. To. If no abnormal value is detected, it is estimated that the user is in an appropriate state for measuring the pulse wave signal and that the PPG sensor unit 122 is in an appropriate state when measuring the pulse wave signal. Will be done. Therefore, in this method, the reliability _ri is calculated based on the ratio of the period during which the abnormal value is not detected within the measurement period of the user's pulse wave signal. Therefore, according to this method, the reliability _ri can be calculated in consideration of the state of the user and the state of wearing the PPG sensor unit 122.

For the reliability _ri in a certain measurement period, for example, the measurement time of the time series data of PPI in the measurement period is set as TM, and the total time obtained by obtaining the PPI value in which no abnormal value is detected in the measurement period is _{T. If N} , it can be expressed by the following mathematical formula (36).

For the reliability _ri in a certain measurement period, for example, the total number of data included in the time series data of PPI in the measurement period is M, and the total number of PPI value data in which no abnormal value is detected in the measurement period is N. Then, it can be expressed by the following mathematical formula (37).

Further, in this method, the reliability _ri may be calculated based on the expression pattern of the abnormal value in the acquired time series data of PPI. For example, the reliability ri in a certain measurement period corresponds to each case of the above-mentioned eight abnormal value appearance patterns in the measurement period, where M is the total number of data included in the time series data of _PPI in the measurement period. Assuming that the number of outliers determined to be outliers for each case is Ni ( _i is a natural number of 1 to 8), it can be expressed by the following mathematical formula (38).

In the above equation (38), αi (i is a natural number of 1 to 8) is a weighting coefficient determined for each case of the appearance pattern of an abnormal value. The weighting coefficient αi is based on, for example, the time-series data of the PPI measured in advance and the environmental information at the time of the measurement (presence or absence of external impact, user's operation pattern, wearing state of the wearable device 10, etc.). , Can be determined experimentally. Further, the weighting coefficient αi may be determined by weighting based on the reason for the occurrence of the abnormal value for each expression pattern of the abnormal value.

The above-mentioned calculation method of the reliability _ri is an example, and the calculation method of the reliability _ri according to the present embodiment is not limited to the above-mentioned method. Further, the above-mentioned methods for calculating the reliability _ri can be combined with each other. In the present embodiment, for example, the reliability ri obtained by the above-mentioned fifth method is referred to as the reliability _ri calculated from the acceleration data acquired by the motion sensor unit ₁₂₄ by the above-mentioned first method. It may be integrated to calculate the final confidence. By combining the above-mentioned calculation method of the reliability _ri in this way, the accuracy of the reliability _ri can be further improved.

<4.5 Reliability output>
By presenting the reliability ri calculated as described above to the user, for example, the user can be made to recognize the reliability of the _HRV index presented together. Further, for example, the calculated reliability ri can also be used when controlling the sensor unit 120 of the wearable device 10 or controlling the calculation process of the _HRV index. Therefore, various examples of the output method of the reliability _ri according to the present embodiment will be described below.

(First method)
First, as an example of the output method of the reliability _ri , the first method of presenting the reliability _ri to the user will be described with reference to FIGS. 24 and 25. 24 and 25 are explanatory views illustrating the output of the reliability _ri according to the first method of the present embodiment. According to this method, for example, by presenting a low reliability _ri to the user, the user is guided to an action of keeping a rest and the same posture in an attempt to improve the reliability _ri . It will be. Further, in this method, when the calculated reliability ri is lower than the predetermined value, it is considered that the reliability of the _HRV index or the like presented to the user is lowered, and the reason is given to the user. Alternatively, an alert may be presented asking for an improvement in reliability _ri .

More specifically, in this method, as shown in FIG. 24, when the stress degree calculated based on the _HRV index is displayed to the user, the display form of the stress degree changes according to the reliability ri. Let me. For example, as shown on the left side of FIG. 24, when the reliability _ri is high, the contrast is increased and the stress degree is displayed. On the other hand, as shown on the right side of FIG. 24, when the reliability _ri is low, the contrast is lowered and the stress degree is displayed. By displaying the display with the contrast changed in this way, the user can easily recognize the reliability of the displayed stress level. Further, for example, as shown in FIG. 25, when the time-series data of the HRV index is displayed as a graph toward the user, the time-series data of the _HRV index during the period when the reliability ri is low is displayed in gray out so as not to be conspicuous. Do 500. By doing so, the user pays attention to the HRV index to be paid attention to, and it becomes easy to confirm the user's HRV index.

As a method of presenting the reliability r _i to the user, the display color, brightness, etc. of the reliability r _i may be changed according to the reliability r _i and displayed, and specifically, the reliability r _i may be numerically displayed. May be displayed, and the method is not particularly limited. In this method, by presenting the reliability _ri to the user in this way, the user is guided to an action of keeping the posture at rest and in the same posture in order to improve the reliability _ri . Will be. Further, the user can easily recognize the reliability of the _HRV index presented together with the reliability ri.

Further, in the present method, when the reliability _ri is presented, the user may be presented with the reason why the reliability of the pulse wave signal measurement is lowered and the method for improving the reliability _ri . In this case, the cause of the decrease in the reliability r _i is estimated by referring to the values of the rest score Sr and the posture score Sp used when calculating the reliability r _i , and the reliability r is based on the estimation result. The reason why _i has decreased and the method for improving the reliability r _i are presented to the user. More specifically, when it is presumed that the user has made a sudden exercise as the cause of the decrease in the reliability _ri , the user is told that "the reliability has decreased due to the abrupt exercise". I will present the presumed reason. Further, in this method, in such a case, a method for improving the reliability _ri such as "Please rest" or "Please stabilize the posture" is presented to the user. If it is presumed that the PPG sensor unit 122 is not attached properly as the cause of the decrease in reliability _ri , the user may be asked to "reattach the wearable device" or the like. , A method for improving the mounting state of the PPG sensor unit 122 will be presented. By making such a presentation, the user can recognize the reason why the reliability _ri is low, so that the user can get a sense of conviction as compared with the case where only the reliability _ri is presented to the user. can. Furthermore, the user can easily recognize the method for improving the reliability _ri and be guided to the execution of the method, so that the accuracy of the pulse wave signal measurement can be improved and the accuracy of the HRV index can be improved. ..

In this method, the reason why the reliability _ri is lowered and the method of presenting the method for improving the reliability _ri are not limited to displaying by characters such as the above-mentioned words. For example, the wearable device 10 may vibrate or the LED provided on the wearable device 10 may emit light based on a pattern associated with each reason in advance, thereby presenting the reason to the user.

(Second method)
Next, as an example of the output method of the reliability _ri , a second method of feeding back the reliability ri to the _wearable device 10 to control the wearable device 10 will be described. In this method, when the period in which the reliability _ri is lower than the predetermined value continues for a predetermined time (for example, several seconds) or more, the operation of the PPG sensor unit 122 for acquiring the pulse wave signal is temporarily stopped. Let me. For example, if the non-resting state continues due to the user running and moving, the above-mentioned rest score Sr and the reliability ri calculated based on the rest score _Sr decrease. If such a decrease in reliability _ri continues for a predetermined time or longer, it is estimated that the low reliability _ri will be maintained even after that, and the operation of the PPG sensor unit 122 (for example, the light receiving operation) is temporarily suspended. Stops the acquisition of pulse wave signals. This is because even if the HRV index is calculated based on the pulse wave signal in a state where the reliability _ri is low, an index that appropriately reflects the user's state cannot be obtained. Further, by stopping the operation of the PPG sensor unit 122, the battery consumption in the wearable device 10 can be suppressed. That is, according to this method, the measurement operation of the wearable device 10 can be brought into a suitable state by feeding back the reliability ri to the _wearable device 10. When stopping the operation of the PPG sensor unit 122, it is preferable to indicate to the user to that effect. Further, the present embodiment is not limited to stopping the operation of the PPG sensor unit 122, and other operations such as communication between the wearable device 10 and the server 30 may be stopped.

Further, in this method, the reliability ri may be fed back to the processing of the detection unit 332, the correction unit 336, and the _HRV index calculation unit 342 of the server 30. Specifically, when the reliability _ri is lower than the predetermined value for a predetermined time (for example, several seconds) or longer, the abnormal value is appropriately detected, the abnormal value is appropriately corrected, and further. It is determined that it is difficult to calculate an appropriate HRV index. Therefore, in this method, in such a case, the detection of the abnormal value by the detection unit 332, the correction of the abnormal value by the correction unit 336, and the calculation of the HRV index by the HRV index calculation unit 342 are temporarily stopped. Let me. As a result, by stopping the operations of the detection unit 332, the correction unit 336, and the HRV index calculation unit 342, it is possible to suppress an increase in the processing resources (processing amount, retained memory amount, stored data amount, etc.) in the server 30. .. That is, according to this method, the processing operation of the server 30 can be brought into a suitable state by feeding back the reliability _ri to the server 30.

Further, in this method, the reliability ri may be fed back to the _HRV index calculation unit 342 of the server 30 to control the HRV index calculation process. Specifically, in this method, the time series data of the PPI acquired during the period when the reliability ri is lower than the predetermined value is excluded from the data when calculating the _HRV index. By doing so, a more reliable HRV index can be obtained. For example, there is a case where it is desired to use the average value of the HRV index calculated in the last month for a certain user for the physical condition management of the user. In such a case, the average value of the HRV index that more accurately reflects the user's condition is calculated by excluding the time series data of the PPI during the period when the reliability _ri is low and calculating the average value of the HRV index. Can be obtained and can be effectively used for user's physical condition management.

Further, in the present method, in the case of integrating the HRV indexes of a plurality of periods to calculate a predetermined index, after weighting each _HRV index based on the reliability ri of each period, the above-mentioned method is performed. The integrated value may be calculated. More specifically, the stress level of a user in a day may be defined by a weighted average of a plurality of SDNNs (a type of HRV index) calculated during the day. In this method, in such a case, a weighting process is performed so that a large weight is given to the _SDNN value in the section having a high reliability ri and a small weight is given to the _SDNN in the section having a low reliability ri. conduct. By performing such weighting, it is possible to suppress the influence of _SDNN in the section with low reliability ri for which the pulse wave signal could not be measured appropriately on the above stress degree, and obtain a highly reliable stress degree. be able to. That is, according to this method, by feeding back the reliability ri to the calculation process of the server 30, the accuracy of the _HRV index, the stress degree, etc. calculated by the server 30 can be further improved.

<< 5. Summary >>
As described above, according to the present embodiment, the reliability of the measurement of the pulse wave signal is calculated, and the reliability is presented to the user together with the HRV index and the like, so that the user's condition is ideally measured. Can lead to a state of affairs. Further, by feeding back the calculated reliability to the control of the wearable device 10 and the server 30, the wearable device 10 and the server 30 are preferably used for calculating the HRV index with high accuracy and suppressing the battery consumption. Can be controlled. That is, according to the present embodiment, the measurement can be set to a suitable state in the measurement of the pulse wave signal in the user under free action.

Furthermore, according to the present embodiment, the HRV index or the like calculated by accurately detecting an abnormal value from the time-series data of the PPI and correcting the time-series data of the PPI based on the detected abnormal value is obtained. , It is possible to avoid deviation from the correct HRV index or the like that should be calculated. That is, according to this embodiment, it is possible to improve the accuracy of the HRV index or the like calculated based on the time series data of PPI.

<< 6. About hardware configuration >>
FIG. 26 is an explanatory diagram showing an example of the hardware configuration of the information processing apparatus 900 according to the present embodiment. In FIG. 26, the information processing apparatus 900 shows an example of the hardware configuration of the server 30 described above.

The information processing apparatus 900 includes, for example, a CPU 950, a ROM 952, a RAM 954, a recording medium 956, an input / output interface 958, and an operation input device 960. Further, the information processing apparatus 900 has a display device 962, a communication interface 968, and a sensor 980. Further, the information processing apparatus 900 connects each component with, for example, a bus 970 as a data transmission path.

(CPU950)
The CPU 950 is, for example, a control unit (for example, the above-mentioned control unit 130) that is composed of one or more processors composed of arithmetic circuits such as a CPU, various processing circuits, and the like, and controls the entire information processing apparatus 900. Functions as. Specifically, the CPU 950 functions in the information processing apparatus 900, for example, the above-mentioned detection unit 332, detection correction control unit 334, correction unit 336, reliability calculation unit 338, HRV index calculation unit 342, and the like.

(ROM952 and RAM954)
The ROM 952 stores control data such as programs and calculation parameters used by the CPU 950. The RAM 954 temporarily stores, for example, a program executed by the CPU 950.

(Recording medium 956)
The recording medium 956 functions as the above-mentioned storage unit 350, and stores, for example, various data such as data related to the information processing method according to the present embodiment and various applications. Here, examples of the recording medium 956 include a magnetic recording medium such as a hard disk and a non-volatile memory such as a flash memory. Further, the recording medium 956 may be detachable from the information processing apparatus 900.

(I / O interface 958, operation input device 960, and display device 962)
The input / output interface 958 connects, for example, an operation input device 960, a display device 962, or the like. Examples of the input / output interface 958 include a USB (Universal Serial Bus) terminal, a DVI (Digital Visual Interface) terminal, an HDMI (High-Definition Multimedia Interface) terminal, and various processing circuits.

The operation input device 960 functions as, for example, the above-mentioned input unit 300, and is connected to the input / output interface 958 inside the information processing apparatus 900.

The display device 962 functions as, for example, the above-mentioned output unit 310, is provided on the information processing device 900, and is connected to the input / output interface 958 inside the information processing device 900. Examples of the display device 962 include a liquid crystal display and an organic EL display (Organic Electro-Luminescence Display).

The input / output interface 958 can also be connected to an external device such as an external operation input device (for example, a keyboard, a mouse, etc.) or an external display device of the information processing apparatus 900. Further, the input / output interface 958 may be connected to a drive (not shown). The drive is a reader / writer for a removable recording medium such as a magnetic disk, an optical disk, or a semiconductor memory, and is built in or externally attached to the information processing apparatus 900. The drive reads out the information recorded on the mounted removable recording medium and outputs it to the RAM 954. The drive can also write records to a removable recording medium mounted on it.

(Communication interface 968)
The communication interface 968 functions as a communication unit 340 for wirelessly or wiredly communicating with an external device such as a server 30 via, for example, the above-mentioned network 70 (or directly). Here, examples of the communication interface 968 include a communication antenna and an RF (Radio frequency) circuit (wireless communication), an IEEE802.5.1 port and a transmission / reception circuit (wireless communication), an IEEE802.11 port and a transmission / reception circuit (wireless communication). ), LAN (Local Area Network) terminal, transmission / reception circuit (wired communication), and the like.

(Sensor 980)
The sensor 980 functions as a sensor unit 120 of the wearable device 10. Further, the sensor 980 may include various sensors such as a pressure sensor.

The above is an example of the hardware configuration of the information processing apparatus 900. The hardware configuration of the information processing apparatus 900 is not limited to the configuration shown in FIG. 26. In detail, each of the above-mentioned components may be configured by using a general-purpose member, or may be configured by hardware specialized for the function of each component. Such a configuration may be changed as appropriate depending on the technical level at the time of implementation.

For example, the information processing device 900 does not have a communication interface 968 when communicating with an external device or the like via a connected external communication device or when the information processing device 900 is configured to perform processing standalone. May be good. Further, the communication interface 968 may have a configuration capable of communicating with one or more external devices by a plurality of communication methods. Further, the information processing apparatus 900 can be configured without, for example, a recording medium 956, an operation input device 960, a display device 962, or the like.

Further, the information processing device 900 according to the present embodiment may be applied to a system including a plurality of devices, which is premised on connection to a network (or communication between each device), such as cloud computing. good. That is, the information processing device 900 according to the present embodiment described above can be realized as an information processing system that performs processing according to the information processing method according to the present embodiment by, for example, a plurality of devices.

<< 7. Supplement >>
It should be noted that the embodiments of the present disclosure described above may include, for example, a program for making a computer function as an information processing apparatus according to the present embodiment, and a non-temporary tangible medium in which the program is recorded. Further, the program may be distributed via a communication line (including wireless communication) such as the Internet.

In addition, each step in the processing of each of the above-described embodiments does not necessarily have to be processed in the order described. For example, each step may be processed in an appropriately reordered manner. Further, each step may be partially processed in parallel or individually instead of being processed in chronological order. Further, the processing method of each step does not necessarily have to be processed according to the described method, and may be processed by another method, for example, by another functional unit.

Although the preferred embodiments of the present disclosure have been described in detail with reference to the accompanying drawings, the technical scope of the present disclosure is not limited to such examples. It is clear that anyone with ordinary knowledge in the art of the present disclosure may come up with various modifications or amendments within the scope of the technical ideas set forth in the claims. Is, of course, understood to belong to the technical scope of the present disclosure.

In addition, the effects described herein are merely explanatory or exemplary and are not limited. That is, the technique according to the present disclosure may exert other effects apparent to those skilled in the art from the description of the present specification, in addition to or in place of the above effects.

The following configurations also belong to the technical scope of the present disclosure.
(1) The reliability of the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user, or the physical index indicating the physical condition of the user calculated from the pulsation fluctuation data. An information processing device including a reliability calculation unit for calculation and a control unit for controlling various processes based on the calculated reliability.
(2) The reliability according to the above (1), which is calculated based on at least one of the user's state, the pulse wave sensor wearing state, and the acquired pulsation fluctuation data. Information processing device.
(3) The information processing apparatus according to (2) above, wherein the reliability is calculated based on the state of the user acquired by the motion sensor attached to the user.
(4) The information processing apparatus according to (2) above, wherein the reliability is calculated based on the wearing state of the pulse wave sensor acquired by the sensor worn by the user together with the pulse wave sensor.
(5) The information processing apparatus according to (1) above, further comprising a detection unit that detects an abnormal value from the pulsation fluctuation data.
(6) The information processing apparatus according to (5) above, wherein the reliability is calculated based on the detected abnormal value.
(7) The information processing device according to (5) or (6) above, wherein the control unit controls the detection unit based on the calculated reliability.
(8) The detection unit extracts the pulsation fluctuation data in the section of the predetermined length from the pulsation fluctuation data, and compares the pulsation fluctuation data in the extracted section of the predetermined length with the predetermined pattern. The information processing apparatus according to any one of (5) to (7) above, which detects the abnormal value.
(9) The information processing apparatus according to any one of (5) to (8) above, wherein the detection unit changes the parameters used when detecting the abnormal value according to the physical characteristics of the user.
(10) The information processing apparatus according to (5) above, further comprising a correction unit that corrects the pulsation fluctuation data based on the detected abnormal value.
(11) The information processing apparatus according to (10) above, wherein the control unit controls the correction unit based on the calculated reliability.
(12) The information processing apparatus according to (10) or (11) above, wherein the correction unit changes the correction process according to the type of the body index to be calculated.
(13) The correction unit extracts the pulsation fluctuation data in the section of the predetermined length from the pulsation fluctuation data, and recognizes and recognizes the pattern of the pulsation fluctuation data in the extracted section of the predetermined length. The information processing apparatus according to (10) or (11) above, which changes the correction process according to the type of the pattern.
(14) The information processing apparatus according to (10) above, further comprising an index calculation unit for calculating the physical index based on the corrected pulsation fluctuation data.
(15) The information processing apparatus according to (14) above, wherein the control unit controls the index calculation unit based on the calculated reliability.
(16) The information processing apparatus according to (14) above, wherein the index calculation unit weights the pulsation fluctuation data or the body index in each section based on the reliability.
(17) The information processing device according to (1) above, wherein the control unit controls the pulse wave sensor based on the reliability.
(18) The control unit performs detection processing for detecting an abnormal value from the pulsation fluctuation data, correction processing for correcting the pulsation fluctuation data, and pulsation fluctuation performed in the information processing apparatus. The information processing apparatus according to (1) above, which controls at least one of the calculation processes for calculating the physical index from the data.
(19) The reliability of the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user, or the physical index indicating the physical condition of the user calculated from the pulsation fluctuation data. An information processing method including calculating and controlling various processes based on the calculated reliability.
(20) The reliability of the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user, or the physical index indicating the physical condition of the user calculated from the pulsation fluctuation data. A program for realizing a computer with a function to calculate and a function to control various processes based on the calculated reliability.

1 Information processing system 10, 10a Wearable device 12 Band part 14 Pressure sensor part 30 Server 50 User terminal 70 Network 100, 300 Input part 110, 310 Output part 120 Sensor part 122 PPG sensor part 124 Motion sensor part 130, 330 Control part 140 340 Communication unit 150, 350 Storage unit 200 Measurement site 202 Vascular 332 Detection unit 334 Detection correction control unit 336 Correction unit 338 Reliability calculation unit 342 HRV index calculation unit 350 Storage unit 352 DB
500 Gray out display 900 Information processing device 950 CPU
952 ROM
954 RAM
956 Recording medium 958 Input / output interface 960 Operation input device 962 Display device 968 Communication interface 970 Bus 980 Sensor

Claims (16)

Reliance for calculating the reliability of the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user, or the physical index indicating the physical condition of the user calculated from the pulsation fluctuation data. Degree calculation unit and
A control unit that controls various processes based on the calculated reliability,
A detection unit that detects abnormal values from the pulsation fluctuation data,
A correction unit that corrects the pulsation fluctuation data based on the detected abnormal value, and
Equipped with
The correction unit changes the correction process according to the type of the body index to be calculated.
Information processing equipment. The information processing apparatus according to claim 1, wherein the reliability is calculated based on at least one of the user's state, the pulse wave sensor wearing state, and the acquired pulsation fluctuation data. The reliability is calculated based on the state of the user acquired by the motion sensor attached to the user.
The information processing apparatus according to claim 2. The reliability is calculated based on the wearing state of the pulse wave sensor acquired by the sensor worn by the user together with the pulse wave sensor.
The information processing apparatus according to claim 2. The information processing apparatus according to claim 1 , wherein the reliability is calculated based on the detected abnormal value. The information processing device according to claim 1 , wherein the control unit controls the detection unit based on the calculated reliability. The detection unit extracts the pulsation fluctuation data in the section of the predetermined length from the pulsation fluctuation data, and compares the pulsation fluctuation data in the extracted section of the predetermined length with the predetermined pattern. The information processing apparatus according to any one of claims 1 to 6 , which detects an abnormal value. The information processing apparatus according to any one of claims 1 to 7 , wherein the detection unit changes a parameter used for detecting the abnormal value according to the physical characteristics of the user. The information processing device according to claim 1 , wherein the control unit controls the correction unit based on the calculated reliability. The correction unit extracts the pulsation fluctuation data in the section of the predetermined length from the pulsation fluctuation data, recognizes the pattern of the pulsation fluctuation data in the extracted section of the predetermined length, and recognizes the pattern of the recognized pattern. The information processing apparatus according to claim 1 , wherein the correction process is changed according to the type. The information processing apparatus according to claim 1 , further comprising an index calculation unit for calculating the physical index based on the corrected pulsation fluctuation data. The information processing device according to claim 11 , wherein the control unit controls the index calculation unit based on the calculated reliability. The information processing device according to claim 11 , wherein the index calculation unit weights the pulsation fluctuation data or the body index in each section based on the reliability. The information processing device according to any one of claims 1 to 13 , wherein the control unit controls the pulse wave sensor based on the reliability. Information processing equipment
The reliability of the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user or the physical index indicating the physical condition of the user calculated from the pulsation fluctuation data is calculated.
Various processes are controlled based on the calculated reliability .
An abnormal value is detected from the pulsation fluctuation data, and
The pulsation fluctuation data is corrected based on the detected abnormal value.
Including that
In the correction of the pulsation fluctuation data, the correction process is changed according to the type of the physical index to be calculated.
Information processing method. A function to calculate the reliability of the pulsation fluctuation data acquired from the sensing data acquired by the pulse wave sensor attached to the user, or the physical index indicating the physical condition of the user calculated from the pulsation fluctuation data. When,
A function to control various processes based on the calculated reliability, and
The function to detect abnormal values from the pulsation fluctuation data and
A function to correct the pulsation fluctuation data based on the detected abnormal value, and
Is a program to realize on a computer
The correction function changes the correction process according to the type of the body index to be calculated.
program.

Images